During the first six weeks of the COVID-19 outbreak in Norway, approximately 2 % of the population was tested for SARS-CoV-2, 5 % of tests were positive, and 5 870 cases were reported to the NIPH.
The beginning of the COVID-19 epidemic in Norway occurred during the winter season's outbreaks of influenza and common cold. It was therefore necessary to have clear criteria for testing in order not to exceed testing capacity, even though COVID-19 diagnostics were in place early and were quickly scaled up. At the beginning of the outbreak, the strategy was to capture all introductions of COVID-19 to Norway, first and foremost related to foreign travel. The start of the outbreak in Norway coincided with extensive foreign travel activity in the winter holiday weeks, and many of the cases detected were men aged 40–60 who had been on skiing holidays in Italy or Austria.
During the days preceding 12 March, the proportion of cases who became infected in Norway increased and the number of cases with unknown source of infection rose
(14). Norway was considered to have entered a new phase of the pandemic (14). When extensive control measures were introduced on 12 March, the criteria for testing were simultaneously shifted towards picking up cases of severe COVID-19, so that illness after foreign travel was no longer sufficient per se to qualify for testing. The restrictions imposed by the testing criteria may well have resulted in some infected persons not being diagnosed.
From 13 March the age distribution of the reported cases changed, and a higher proportion were hospitalised or died. In our data, hospitalisation was recorded at the time of reporting. The median time interval between onset of symptoms and hospitalisation and/or death indicates that some cases may have been hospitalised after the time of reporting or have died after 7 May. However, the total number of cases hospitalised is consistent with the numbers in the Norwegian Intensive Care and Pandemic Registry
(15). Even though a history of travel was no longer among the prerequisites for testing after 12 March, a large number of cases who had travelled to Spain were reported in the weeks that followed. Many of them were probably tested because of high age and/or increased risk of severe illness, and not primarily because of their travel. This example reveals a bias in registered information about country of exposure, which was attributable to the testing criteria applying at the time.
The proportion of healthcare workers among cases was higher in the latter part of the study period, but this should be interpreted with caution because of incomplete data. Moreover, the type of exposure and nature of occupation (for example patient contact) of most health personnel was unknown. The numbers therefore cannot be used as a measure of how many were exposed at work. More reliable data can probably be acquired by improving the surveillance systems. We also found that after 12 March, the proportion of female cases exceeded that of males, possibly due to increased testing among healthcare workers and the elderly, both groups with a larger proportion of women
The median interval between onset of symptoms and testing increased from 13 March. This may be because the testing targeted patients with severe disease, which takes some time to develop, or it may be a result of local practice and testing capacity. The discrepancy between numbers of positive tests and numbers of reported cases can be explained by the fact that some cases were tested repeatedly and were thus reported to the laboratory surveillance system several times. Furthermore, some cases tested up to 5 April were reported only after that date. However, there was very little delay between testing and reporting throughout the study period. Data reported to the Outbreak Register and MSIS are therefore considered to be a timely source of information on the status of the outbreak, also early in the epidemic. Case data were maximised by linking together data from MSIS and Outbreak registers, while data on testing from the National Reference Laboratory were important for understanding the fluctuations in the positivity rate in light of testing activity.
The number of cases registered in the three data sources will not represent the true number of persons infected with SARS-CoV-2 in the population. Only laboratory-confirmed cases are reported. The possibility that there was infection in Norway before 26 February cannot be ruled out, as a case with symptoms that could be attributable to COVID-19 was observed as early as 12 February, but only reported later. Moreover, the criteria for testing during this period were such that only those with symptoms met the criteria, and a large percentage of COVID-19 cases are asymptomatic, or only have mild symptoms
(17, 18). Large-scale testing of asymptomatic or pre-symptomatic persons would have required significant resources, with increased risk of false positive and false negative test results. Although the discrepancy between number of registered cases and the true incidence of infections in the population is unknown, there is reason to believe that the incidence of infection in Norway during the study period was relatively low compared to many other countries (19). The seroprevalence (proportion of persons with antibodies) in Norway was estimated to be around 1 % in mid-April (20).
Access to timely surveillance data has been crucial for monitoring the epidemic. The compilation of data from various sources has contributed to a more comprehensive picture of the onset and development of the epidemic. Among the cases of COVID-19 reported to the NIPH in the first six weeks of the epidemic, differences were found in the distribution of age, sex, place of exposure and disease severity in the periods before and after 12 March, when extensive control measures were introduced. While some of these differences may reflect different phases of the epidemic, it is important to interpret the findings in the light of criteria for testing, testing activity, control measures and characteristics of the surveillance systems. Additional studies, information from more data sources and further development of surveillance systems are required to provide both broader and more detailed knowledge of the extent and evolution of the epidemic in Norway, where and in what contexts transmission occurs and who is at greatest risk of severe illness. Sound surveillance data which can be linked to various data sources are crucial for following up the current strategy, which is based on early detection of cases and local management by means of targeted measures.